CARL GOTZMER,

Chemistry Research, Naval Ordnance Station, Indian Head, Maryland 20640. Received December 27, 1972. The irradiation of benzophenone dissolved in a ...
1 downloads 0 Views 565KB Size
2964

J . Org. Chem., Vol. 38, No. 17, 1973

GOTZMER, MUELLER,A N D CZIESLA

Tetrafluorohydrazine as Radical Scavenger in the Photoreduction of Benzophenone CARLGOTZMER, JR.*, KURTF. nIlUELLER, AND A i . J. CZIESLA Chemistry Research, Naval Ordnance Station, Indian Head, Maryland

20640

Received December 27, 1972 The irradiation of benzophenone dissolved in a hydrocarbon in the presence of tetrafluorohydrazine was found to yield equimolar amounts of difluoraminodiphenylmethanol and the respective difluoramino-substituted hydrocarbon. The hydrocarbons investigated include cyclohexane, 3-methylpentane, toluene, and cumene. The rates of tetrafluorohydrazine consumption were measured.

The mechanism of the photoreduction of benzophenone in solution has been investigated in many laboratories since the pioneering experiments of Ciamician and Silber.' There is much supporting evidence that the first excited triplet state of benzophenone is the dehydrogenating species in t,hese systems. The formation of t'he products isolated was explained by combination and disproportionation reactions of the monoradicals formed in the dehydrogenation stepa2 More recent'ly this mechanism has been modified by postulation of intermediate adduct's3-b or chargo t'ransfer complexes6 to accommodat,e conflict'ing experimental evidence. Evidence for the existence of the postulat'ed monoradicals by trapping experiments has not been reported t'o our knowledge. The commonly used scavengers for radicals, such as diphenyl picrylhydrazyl or galvinoxyl, cannot be used because they absorb st'rongly in the wavelength region utilized for the photoreduction of aromabic ketones. Superficially, the formation of benzpinacol itself could be considered as such a t'rapping experiment. However, the initial formation of 1,2-diphenyl-1,2bis(p-chlorophenyl)-1,2-ethanediol during irradiation of p-chlorobenzophenone and benzhyd1-01~and the init'ial formation of unlabeled benzpinacol during the photolysis of benzophenone and benzhydrol (labeled with 14C)7in benzene solution indicated that the isolated products do not arise simply by combination reactions of t'he postulated initial monoradical. Also, t'he formation of terebic acid during photolysis of benzophenone in isopropyl alcohol in the presence of maleic anhydride* has been considered as proof for a trapping reaction of the postulated dimethyl hydroxymethyl radicals. However, no 2,2-diphenylparaconic acid, which should form from t'he postulated diphenyl hydroxymethyl radicals in an analogous fashion, could be isolated in t'hesc experiment^.^ The latter reaction has been shown to occur when diphenyl hydroxymethyl radicals are thermally generated in the presence of maleic anhydride.9 More recently, tertnitrosobutane has been employed as a radical scavenger. The rcport'ed results, however, are inconclu(1) G. Ciamioian and P. Silber, Chem. Ber.. 33,2911 (1900). (2) G. Weismann, E. Bergmann, and Y. Hirshberg, J . Amer. Chem. Sac., 60, 1530 (1938). (3) G. 0. Sohenck, W.Meder, and M. Pape, "2nd U N Generva Conference," Pergamon Press, Elmsford, N. Y., 1959, p 352. (4) J. N. Pitts, R. L.Letsinger, R. P. Taylor, J. M.Patterson, G. Recktenwald, and R . B. Martin, J . Amer. Chem. Sac., 81, 1068 (1959). (5) N. Filipescu and F. L. Minn, J . Amer. Chem. Soc..90, 1544 (1968). (6) V. Franzen, Justus Liebigs Ann. Chem., 638, 1(1960). (7) Unpublished results by one of the authors (M. J. Cziesla) in ooopera-

tion with M. Pape (BASF, Ludmigshafen, Germany). (8) G. 0. Sohenok, H. Koltzenburg, and H. Grossmann. Angew. Chem., 69, 177 (1957). (9) G. 0 . Schenok, G. Mathias, M. Pape, M. Cziesla. and G. Von Buenau, Justus Liebigs Ann. Chem., 719, 80 (1968).

sive. Whereas Leaver and Ramsay 'O reported the trapping of dimethyl hydroxymethyl radicals by tertnitrosobutane during the photoreduction of benzophenone in isopropyl alcohol, Perkins and Ward l' reported the formation of tert-butyl nitroxide in the reaction of tert-butyl peroxyoxalate and isopropyl alcohol in the presence of tert-nitrosobutane. They attribute the formation of tert-butyl nitroxide to a hydrogen transfer reaction from the dimethyl hydroxymethyl radical to tert-nitrosobutane. Leaver and Ramsay also observed the tert-butyl nitroxide; however, they postulate its formation to arise from a hydrogen transfer from the diphenyl hydroxymethyl radical. I t is concluded that trapping experiments in radical reactions are significant evidence for the proposed mechanism only if all postulated radicals are trapped simultaneously and quantitatively. Tetrafluorohydrazine mas expected to be an effective radical scavenger in photochemical dehydrogenation reactions because of its reported properties and reactions. It has been shown to be in equilibrium with difluoroamino radicals a t ambient temperature12 and to react with monoradicals produced by either FzNNFz

2.NF2

thermal or p h o t o ~ h e m i c a l ~decomposition. ~ Undesirable side reactions were not expected, since no ionic addition reactions of tetrafluorohydrazine have been reported and the difluoroamino radicals have been found to abstract hydrogen only at elevated temperatures.I4 It has been reported in a previous publication15 that a-difluoramino ethers were formed by irradiation of an ether solution of benzophenone in the presence of tetrafluorohydrazine, This reaction was extended to solutions of benzophenone in hydrocarbons and a detailed investigation of all reaction products was performed. Results Difluoraminoalkanes. -Benzophenone dissolved in hydrocarbons was irradiated with a high-pressure mercury lamp while purging with tetrafluorohydrazine from a reservoir. Figure 1 represents a typical plot of gas consumption us. time. The rate of tetrafluorohydrazine uptake slowed down during irradiation I. H. Leaver and A. Caird Ramsay, Tetrahedron, 25, 5669 (1969). M .J. Perkins and P. Ward, J . Chem. Sac. B , 395 (1970). C. B. Colburn and A. Kennedy, J . Amer. Chem. Sac., 80, 5004 (1958). R. C. Petry and J. P. Freeman, J . Amer. Chem. Sac., 83, 3912 (1961). J. Grzechomiok, J. F. Kerr, and A. F. Trotman-Diokenson, Chem. Cammun., 109 (1965). (15) M. J. Cziesla, K. F. Mueller, and 0. Jones, Tetrahedron L d l . , 813 (10) (11) (12) (13) (14)

(1966).

J. Org. Chem., Vol. 38, No. 17, 1973

PHOTOREDUCTION OF BENZOPHENONE and stopped after an amount of gas approximately equimolar to the starting benzophenone was consumed. During irradiation the solution turned yellow; however, this color was not connected with the decrease in tctrafluorohydrazine consumption. The addition of more benzophenone to the yellow reaction solution again started the photoreaction with tetrafluorohydrazine a t the same rate as a t the beginning of the experiment. I n Table I, the tctrafluorohydrazine consumption rates for the hydrocarbons investigated are listed.

500

-u3'

400

-

3cQ

-

-

z"

-

m-

-

TABLE I CONSUMPTION OF TETRAFLUOROHYDRAZINE AND RELATIVE QUANTUM YIELDS Substrate

Cumene Cyclohexane 3-Methylpentane Diethyl ether (15) Dioxane (15) Tetrahydrofuran (15) Diisopropyl ether (1.5)

Consumption, ml/rnin

Re1 6

3.58 3.5

1.02 1

2.38 3.09 2.25 1.67 0.80

0.68 0.88 0.64 0.48 0.23

These values were taken from the initial straight line of plots as in lcigurc 1. The hydrocarbon solvents and volatile reaction products were distilled together after irradiation. Vapor phase chromatographic analysis of the distillates indicated one photoproduct in the experiments with toluene and cyclohexane and one major and several minor products in the experiments with 3-methylpentane. The sum of the minor products, however, was less than 5% of the main product. No efforts were undertaken to elucidate the structures of these minor products. T h r other photoproducts were identified as a-difluoroaminotoluenc, difluoroaminocyclohexane, and 3-difluoroamino-3-methylpent ane, respectively . Two products were observed in the distillates of the cumenc experiments; infrared analysis indicated the presence of difluoroamino groups (absorption a t 885 cm-l). During attempted separation and purification of the two products by distillation, one of them disappeared accompanied by a corresponding increase in the concentration of the other product. The latter showed no absorption at 885 cm-l and was identified as a-methylstyrene, indicating that the product which disappeared was indeed the expected cydifluoraminocumenc which lost difluoraminc during attempted distillation.

2965

1

0

2

3

Photolysis Ihwrsl

Figure 1.-Photolysis of 4 g of benzophenone in 400 ml of cyclohexane in the presence of tetrafluorohydrazine.

The benzophenone-cyclohexane solution was photolyzed with tetrafluorohydrazine, purging until the gas consumption had ceascd; 0.9 mol of tctrafluorohydrazine was consumed for every 1 mol of starting benzophenone. Then the solvent and the volatile reaction product were removed under vacuum at room temperature. The viscous distillation residuc exhibited hydroxyl, difluoramino, and phenyl absorption in the infrared, but only a trace of a carbonyl band could be observed. After hydrolysis of the residue with aqueous acetic acid, over 96% of t he original benzophcnone could be isolated. Elemental iodinc was liberated when the hydrolysis was carried out in the presence of potassium iodide; titration of the iodinc gavc valucs of 1.8 molecules of iodine liberated for each molecule of starting benzophenone. All available data are consistent with the structure of difluoraminodiphenylmethanol for the residue. Compounds containing a difluoramino and a hydroxyl group a t the same carbon are known16 to be easily hydrolyzed to the parent carbonyl compound and diPh

Ph

NF2--d--oH

I

Ph

-+ c=o I I

+ HNF~

Ph

fluoramine. Difluoraminc itself reacts quantitatively with potassium iodide in acidic solution. l7 HNFI

+ 4HI +212 + NHiF + HF Discussion

Difluoraminodiphenylmethano1.-Material balance of the reactions indicated that only about one half of the consumed tetrafluorohydrazine could be accounted for in the isolated difluoraminoalkanes. The reaction with cyclohexane \vas selected for a careful investigation of the distillation residue.

Mechanism.-The isolation of equimolar amounts of difluoraminocyclohexane and difluoraminodiphenylmethanol constitutes the first example in which both monoradicals, postulated as intermediates in the photo(16) (a) J. P. Freeman, W. H. Graham, and C. 0. Parker, J. Amer. Chem. Soe., 90, 121 (1968); (b) 9. F. Reed, Jr., J. Org. Chem., S2, 2894 (1967). (17) E. A. Laivton and J. Q.Weber, J. A n e r . Cham. Soc., 81,4755 (1959).

2966 J . Org. Chem., Vol. 38, No. 17, 197.9

GOTZMER, MUELLER,AND CZIESLA

reduction of benzophenone (steps 1 and 2 below), were trapped simultaneously and quantitatively.

-%lPhzCO +3PhzC0 3PhzC0 + RH +PhzCOH + R . PhzCO

+ *NFz--+ PhzC(0H)NFz R . + *NFz+RNFz PhzC(0H)NFz + +NF2 PhzCOH + FzNNFz R * + FzNNFz +RNFz + -NFz PhzCOH

--f

(1) (2)

(31 (4) (5)

Substrate

This paper

Ref 19

Cyclohexane Cumene 3-Methylpentane 2,3-Dimethylbutane Diethyl ether Dioxane Tetrahydrofuran Diisopropyl ether

1 12.2 8.2

1 19.8

6.0

2.6 1.0 1.4 1.4

(6)

The reported experiments do not yield conclusive evidence that the monoradicals originating according to reaction 2 form the difluoramino compounds exclusively via reactions 3 and 4. Another attractive possibility is their formation through reaction of the monoradicals with undissociated tetrafluorohydrazine according to 5 and 6. The low bond energy of the N-N bond in the tetrafluorohydrazine of 19.9 kcall8 makes reactions 5 and 6 exothermic by about 50 kcal. Considerations with respect to the concentration of the different radicals in the solution during photolysis further support the possibility of reactions 5 and 6. High efficiency of the trapping reactions 3-6 was demonstrated by the fact that neither benzpinacol nor mixed radical combination products could be isolated in the experiments carried out in the presence of tetrafluorohydrazine. Benzpinacol would have precipitated during photolysis in the experiments with cyclohexane and 3-methylpentane, since it is practically insoluble in the hydrocarbons. I n the system benzophenone-cyclohexane, 96.3% benzophenone was recovered of the hydrolysis of the distilration residue. Since the theoretical mixed combination product would not yield benzophenone under the reaction conditions, cage combination reactions did occur to less than 4y0if at all. Relative Quantum Yield for Hydrogen Abstraction by Triplet Benzophenone.-The observed rates of tetrafluorohydrazine consumption in the different hydrocarbons should be directly proportional to the quantum yield for hydrogen abstraction by triplet benzophenone providing that the quenching of the excited states of benzophenone by tetrafluorohydrazine is negligible. Whether or not this assumption is correct will be investigated in future experiments. Preliminary data do not indicate a significant quenching. The measured consumption rates of tetrafluorohydrazine are presented in Table I together with values from previous experiments with ethers. l5 The differences in relative quantum yields can be rationalized by a combination of electronic and steric effects. Selectivity of Triplet Benzophenone.-In the experiments with 3-methylpentane more than 95% of the reaction product resulted from the attack of the triplet benzophenone on the tertiary hydrogen, indicating high selectivity of the excited benzophenone. Similar results were reported in experiments with ethers, where only products could be isolated which resulted from hydrogen abstraction adjacent to the ether oxygen. In Table I1 relative reactivities per hydrogen are tabulated together with results reported by Walling (18) J. A. Kerr, Chem. Rev., 66,496 (1966).

TABLE I1 RELATIVEREACTIVITIES PER H ATOM

and Gibian.l9 The values from the experiments with triplet benzophenone are in fair agreement. They show the same high selectivity for the hydrocarbons. Conclusion The results reported in this paper have supported our original postulation16 that tetrafluorohydrazine can be used as a valuable “reagent” in the investigation of the mechanisms of photochemical reactions and pinpoint the site of attack of excited states. Experimental Section General.-Infrared and ultraviolet spectra were obtained with Beckman IR-8 and Bausch & Lomb 505 spectrometers, respectively. The nmr spectra were obtained with a Varian D P 60 spectrometer using deuteriochloroform as the solvent and tetramethylsilane as an internal standard. Reaction mixtures and distillation fractions were monitored on a Perkin-Elmer 154 gas chromatograph and the preparative gas chromatograph work was performed on a Beckman Megachrom equipped with standard Beckman 6 ft X 0.75 in., 20% Paraplex G-25 on 42-60 Chromosorb W columns. Analyses were performed by Galbraith Laboratories, Inc., Knoxville, Tenn. Materials.-Tetrafluorohydrazine (Air Products, 99% pure) was used as received. Toluene (Matheson Coleman and Bell, spectroquality), benzophenone (Fisher, reagent grade), isopropyl alcohol (Fisher, spectrograde), cumene (Baker, Baker grade), and 3-methylpentane (Aldrich, research grade) were used without further purification. Apparatus and Procedure.-All irradiations were performed in the following manner. Solutions consisting of 4.0 g of benzophenone in 400 ml of the appropriate hydrocarbon solvent were placed in a photoreactor which was equipped with a fritted gas inlet tube at the bottom and gas outlet tube at the top. A jacketed, water-cooled, 100-W Hanovia high-pressure mercury lamp (Type SOL, 608-368) was immersed in the reactor utilizing a Pyrex 7740 filter to protect the difluoramino radicals and the expected products from the light. Tetrafluorohydrazine was pumped through the solution with a Cole-Parmer Masterflex tubing pump during irradiation and recycled to a reservoir to measure gas consumption. The solutions were thoroughly purged with purified nitrogen both prior to introduction of tetrafluorohydrazine and after completion of irradiation to avoid potentially hazardous mixtures of organic compounds, tetrafluorohydrazine, and oxygen. None of the reported photochemical reactions occurred in the absence of ultraviolet light. The difluoramino compounds and unreacted solvents were separated from the reaction mixture by distillation under vacuum followed by preparative glpc to give the pure difluoramino compounds. Photolysis of Toluene Solution.-Tetrafluorohydrazine was bubbled through a solution of benzophenone in toluene and irradiated at ambient temperature for 3 hr, yieldingonlyonephotoproduct that could be detected by glpc analysis. The unreacted toluene was removed by vacuum distillation (80 mm, 50’). The remaining residue was vacuum distilled a t 1 mm and 50’ to c01lect the photoproduct and some toluene. The distillate, containing photoproduct and toluene, was separated by preparative glpc to give the analytically pure sample of or-difluorammo(19) C. Walling and M. J. Clibian, J . Aner. Chem. Soo., 87,3361 (1966).

J. Org. Chem., Vol. 38, No. 17, i973 2967

PYROLYSIS OF SULFIDE TOSYLHYDRAZONE SALTS toluene: ir (neat) 805,890, and 925 cm-l (-NF2); nmr (CDCl,) 6 4.58 (t, 2, J = 29 Ha, -CHzNF,) and 7.37 ppm (s, 5, C6H5-). Anal. Calcd for C~HTFZN:C, 58.74; H, 4.93; F, 26.54; N , 9.78. Found: C, 58.92; H , 5.19; F, 26.58; N, 9.59. Photolysis of Cumene Solution.-The above photolysis prophotocedure was repeatedsubstituting cumene for toluene. TWO products were detected by glpc analysis. The unreacted cumene was removed by vacuum distillation (15 mm, 50') and the remaining residue was distilled a t 1 mm and 50'. After the distillate was purified by preparative glpc only one of the original photoproducts could be detected. The photoproduct, that decomposed showed difluoramino absorption at 885 cm-l and the remaining photoproduct was passed through the preparative glpc to give an analytical sample of a-methylstyrene. Anal. Calcd for CgHlo: C, 91.46; H , 8.54. Found: C, 91.38; H, 8.51. Photolysis of 3-Methylpentane Solution .-The photolysis procedure was repeated using 3-methylpentane as the solvent with an irradiation time of 4 hr. One major and several minor (total